Research >> 1. Energy Saving from HVAC System >>■ Automatic commissioning method for HVAC system based on mathematical models


■ Automatic Commissioning Method for HVAC System Based on
 Mathematical Models


□ Fault Detection and Diagnosis Method for VAV Terminal Units

This research proposes two fault detection and diagnosis methods for VAV units without a sensor of supply air volume, and the results of applying these methods to a real building are presented.
 One method detects faults by applying a statistical method to four values calculated using the room air temperatures and the demand values of VAV damper opening of each unit during a steady state operation period. From the results of case studies, the method can reduce the number of units to be checked as faulty ones down to 12% of the total number and all the units that really have a fault are included in this group.
 The other method judges the faults by applying dynamic system analysis to the operational data when the VAV system starts up. From the result of the case studies, the method can reduce the number of units down to 30%, among which five units actually have a fault and only one faulty unit was not included in this group. Both methods can reduce time and cost for commissioning of VAV units significantly by the help of BEMS.



Figure-1 Flow of Grubbs' test


Figure-2 Result of the test


□ Estimation of Excessive HVAC Energy Consumption due to Faulty VAV Units

The research investigates, both experimentally and by simulation, the influence of faulty VAV units that do not affect the zone temperature on total energy consumption of a VAV system. In an experimental room, the respective total energy consumptions of both normal operation and faulty operation were measured.
 The results indicate that a faulty unit affects other normal units, causing an energy waste of approximately 20-50%. In addition, the present paper also discusses whether the additional energy consumption can be calculated via the simulation of the VAV system in the experimental room. The results indicate that the simulation can be used to estimate the additional energy consumption caused by the faults.



Figure-3 Experimental room


Figure-4 Simulation model by MATLAB/Simulink